B-162 Evaluation of two commercially available stable isotope labeled whole-protein infliximab as internal standard for a clinical infliximab LC-MS/MS assay

Abstract

Abstract Background Liquid chromatography tandem mass spectrometry (LC-MS/MS) based quantitation of infliximab has been successfully introduced in the clinical laboratory for therapeutic drug monitoring in recent years. Although stable isotope labeled (SIL)-peptide or non-labeled analog protein internal standards have been utilized, SIL-whole protein infliximab is generally considered to be the ideal internal standard for the correction of variations during sample preparation and analysis. In this study, we compared two commercially available SIL infliximab proteins, Sigma SILu™MAb and Promise Proteomics, for use in a clinical infliximab assay. Methods Both labeled infliximab proteins incorporate [13C6, 15N4]-arginine and [13C6, 15N2]-lysine in their sequence and are designed for use as an internal standard for LC-MS/MS based quantitative analysis of infliximab. For the comparison of their performance, a laboratory-developed LC-MS/MS clinical method was utilized. Linearity at 6 levels in triplicate with Remicade and 3 biosimilars (-abda, -axxq, and -dyyb), intra-assay precision at 2 levels (n=10/level), and method comparison (n=15) experiments were performed using each labeled infliximab as the internal standard. For sample preparation, internal standard was mixed with the serum sample, followed by protein precipitation with ammonium sulfate, protein denaturation, and tryptic digestion. The peptides YASESMSGIPSR (YASE) from infliximab and YASESMSGIPSR [13C6, 15N4] (SIL-YASE) from labeled infliximab were monitored by LC-MS/MS (Transcend TLX and TSQ Atlis, Thermo Scientific) and utilized for the quantitation of infliximab. Results For the linearity experiment (1-50 µg/mL), while utilizing the Promise Proteomics internal standard to assay had an accuracy range of 103.4% to 114.1% for Remicade, 100.6% to 108.2% for Infliximab-axxq, 103.0% to 110.9% for Infliximab-abda, 102.7% to 108.3% for Infliximab-dyyb. For the Sigma SILu™Mab SIL infliximab, the accuracy range was 92.1% to 99.6% for Remicade, 94.5% to 107.2% for Infliximab-axxq, 99.0% to 104.2% for Infliximab-abda, and 105.1% to 112.0% for Infliximab-dyyb. The mean overall error for Remicade and biosimilars was 0.203 µg/mL (2.0%) for Sigma SILu™Mab SIL infliximab and 0.193 µg/mL (2.0%) for Promise Proteomics internal standardization. Intra-day precision using each SIL infliximab was evaluated with Remicade at low and high levels (5 and 20 µg/mL). Using Promise Proteomics, the coefficient of variation (CV) was 2.8 and 2.2%, while using Sigma SILu™Mab the precision was 6.8 and 6.5% for the low and high levels, respectively. The regression analysis for the method comparison experiment had a correlation coefficient (r) of 0.9996 with a slope of 1.050 (95% confidence interval, 1.030-1.070) and intercept of -0.318 (95% confidence interval, -0.842 to 0.205). Conclusions The mean overall error for the linearity experiment was the same using Sigma SILu™Mab or Promise Proteomics as the internal standard. The Promise Proteomics product demonstrated a lower intra-assay CV compared to the Sigma SILu™Mab. The 2 products displayed a strong correlation during the method comparison experiment. Both Sigma SILu™Mab and Promise Proteomics labeled standards showed acceptable performance as an internal standard for LC-MS/MS based quantitative analysis of infliximab. Having both products available increases the robustness of LC-MS/MS clinical assays in the event of supply issues.